Hybrid approach to Non-Replicated Dynamic Fragment Allocation in Distributed Database Systems

Abstract

Distributed Database has been used as the most essential technology for business organizations in recent years. The data allocation is a critical issue in distributed database design. Allocation of data among different sites across the network effects on the performance of the distributed database system. The static allocation supports only limited response to the change in workload. This paper proposes a hybrid strategy for non-replicated dynamic fragments allocation in distributed database system. The proposed hybrid algorithm is modified by using selective concepts of Threshold Time Volume and Distance Constraints algorithm and Region based Fragment Allocation algorithm. This algorithm modifies Threshold Time Volume and Distance Constraints algorithm by considering a concept of region instead of distance factors. It reallocates fragment with respect to the average maximum data volume of that region access to fragment instead of considering the high frequency of individual site. It will reduce the migration of the fragment over different sites and also improve the overall performance of the system.